Milton Piketty

I can see that the search for constants may be a valuable way to guide research. Clearly there are empirical regularities in economics that speak of the existence of constants. I have in mind such phenomena as the extraordinary similarity of long-term interest rates over centuries, the stability of the relationship between the velocity of money and interest rates, the similarity of lag patterns around the world. There is no reason in principle why there should not be constants underlying that kind of behavior even though they do deal with human action and reaction.--Milton Friedman to Eric Schliesser, August 24, 2004.

 

In any case, this virtual stability of the pure return on capital over the very long run (or more likely this slight decrease of about one-quarter to one-fifth, from 4-5 percent in the eighteenth and nineteenth centuries to 3-4 percent today) is a fact of major importance for this study.--Thomas Piketty, Capital,  p. 206; [see this chart]

 

In a country, such as Great Britain, where money is lent to government at three per cent and to private people upon a good security at four and four and a half, the present legal rate, five per cent, is perhaps as proper as any.--Adam Smith, Wealth of Nations, 2.4

After dipping in Piketty's fashionable book, I have started to read it with a great deal of care. I look forward to posting about it regularly over the next few months.

 

Piketty is the true heir of Milton Friedman. This claim might seem perverse if one focuses on policy. But if one looks at (a) methodology, and, crucially, (b) the conception of what economics might be about, ultimately, then Piketty's book is an attempt to return economics to an approach that was never really dominant, but that can be book-ended between Adam Smith's Digression on Silver (or Hume's population essay) and Friedman's (1963) Monetary History. This approach to economics is focused on (i) the longue duree (centuries rather than quarters); (ii) data driven while being extremely sensitive to limitations in -- and self-conscious about the social construction of -- one's data; (iii) a suspicion of intricate mathematical modeling accompanied by a privileging of simple models; (iv) resolutely macro; (v) a recognition of the political (legal, etc.) context of economic activity. There is also a more subtle similarity: (vi) the aim of economics is to generate an "intellectual climate" that can influence policy (Piketty, 549). Now, in his book Piketty gives (i)-(v) a name, "serial history." This focuses on "economic and social history based on the evolution of prices, incomes, and fortunes." (p. 582 n. 35)  And, as he admits at the end of his book, the proper name for (i)-(vi) is: "political economy." (574)

Now, the previous paragraph may seem fanciful. There is little evidence in Piketty's book that he has read Hume or Smith. At the start of his book, Piketty briefly mentions Malthus (and Ricardo), but he does not have genuine working knowledge of them;* the only pre-twentieth economist that he mentions with any insight is Marx (who, of course, really knew his classical economics). Even so, it is less silly than it may seem, if one realizes that a crucial figure in Piketty's narrative is Kuznets (the third Nobel laureate in economics). This is so because part of Piketty's empirical argument is to refute the so-called Kuznets curve. The larger significance of Kuznets is that he got national income accounting off the ground and generate standardized measurements of GDP. As Piketty puts it: "Kuznets's work in establishing the first US national accounts data and the first historical series of inequality measures was of the utmost importance and it is clear from reading his books (as opposed to his papers) that he shared the true scientific ethic." (14)

 

The reason why Kuznets is "of the utmost importance" to Piketty is that, by all accounts, he was the greatest practitioner of "serial history." Kuznets' most gifted student was Milton Friedman. (Kuznets and Friedman wrote a once famous book together.) And Friedman's own greatest contribution to serial history is his Monetary History (co-written with Anna Schwartz), which "follows in minute detail the changes in United States monetary policy from 1857 to 1960." Piketty labels it a "monumental" and "fundamental work." (548) It is, I think, the only book called 'fundamental' (which, in Piketty's vocabulary has a very favorable connotation--he loves 'fundamental' laws, for example); there are only two other works labeled, "monumental," these are written by Kuznets and Kuczynski (both admired by Piketty). 

 

That is to say, the methodological significance of Piketty's project is, that it is self-consciously designed to revive "serial history:" its "decline is unfortunate as well as reversible." (582 n 35; hence 'Milton Piketty.') One reason why it is reversible is that computing power is now cheap (so one can handle huge amount of data) and that one can make it more exciting by relegate the technical appendices to "spreadsheets and online databases." (577)

 

Now, just to show my own hand, in  my view serial history is of utmost methodological importance. (I will  return to this in future posts.) Serial history provides the only hope, thus far, for economics to move from (i) empirical regularities to unearth (ii) the existence of constants. Until I read Piketty's book, which (as the quote at the top of this post suggests), has a considerable interest in establishing and exploiting the significance of empirical regularities, the one (and so far only) major economist of the last half century that really grasped this point is Milton Friedman (see above; Friedman is responding to my paper). For without those constants (and the mechanisms by which they are reinforced and operate), we will never understand the deep institutional  and psychological causes of political economy. In fact, Piketty's criticism of Friedman's Monetary History, offered en passant, is from this vantage point also a deep one: it "says nothing about the virtues and limitations of other institutions." (549)

 

Now, one might wonder why Piketty has not made these threads more transparent. I think part of the explanation is that he is fully aware that Friedman subsequently became (legitimately) understood as fostering a "suspicion of the ever-expanding state" (549). That is, Friedman's political aims overshadowed his (largely discarded) methods (I have told this story here). But the other part of the explanation is that Piketty can make a very similar point in a French context. For, he singles out François Furet's "intellectual trajectory" to explain the demise of serial history as a "research program" (575). (Furet lived long enough so that I could briefly overlap with him at The University of Chicago.) To praise Furet as a serial historian at the end of a book, which attacks Bourgeois-Liberal (and Libertarian) ideological complacency and ends with advocacy of a global tax on capital, is just as ironic in a French context as it would have been to praise Friedman Stateside.**

 

 

*Piketty really seems to have learned his craft from two important collaborators: Anthony Atkinson and Emmanuel Saez.)

 **To be clear: Furet would have enjoyed it.

Comments

[Mark Anderson]

Precisely the sort of writing--historical-explanatory--for which I have long appreciated your posts (especially those in the history of philosophy and science). Thank you, and I look forward to your future pieces on this book.

[Anna Alexandrova]

Thank you for this, Eric! Very revealing. Two questions that are both very big and you should feel free to ignore them:
A) do you think Friedman's Monetary History is at odds with his Methodology of Positive Economics? (You probably mentioned it before but I forget). In which case it's not all the way Milton Piketty, right?
B) how does serial history compare to the sort of macro history that we encounter in Jared Diamond?

Looking forward to your entries as you read...

[eric Schliesser]

A) Kevin Hoover has done interesting work on this question, so you should really ask him. From my end: it depends a lot on what you take F1953 to be saying (and that is very controversial). If you think it is about a defense of instrumentalism then there is a serious tension with the Monetary history. But I think that's just another reason to reject the instrumentalist reading of F1953.
B) I think serial history has relatively standardized and stabilized (constructed) data available that run long periods. I don't think that's available to us in natural history; I tend to think of Diamond's work more about marshaling, very creatively converging evidence. But there is no doubt that serial history has commonalities with the kind of Darwinian/geological/anthropological projects of Diamond. What do you think?

[anna alexandrova]

OK thanks. On (A): It's not like F1953 with a realist reading is any more friendly to the sort of data-driven approach that serial history exemplifies. Minimal models even when interpreted in a realist fashion (i.e. in a Galilean way) are still minimal. So either way we have a tension.

[eric Schliesser]

Ah, we disagree, Anna! But thank you for pushing the point. For, it is the minimal model that tells you what data really counts as data and what counts as noise. In Monetery history, Friedman has a fairly minimal (Monetarist) model and he uses it to interpret and test the data. (He also finds an alternative model wanting in light of the data.) One way to understand Piketty's criticism is that in the monetary history the model is too minimal that Friedman fails to explore alternative institutions.

[anna alexandrova]

i see, perhaps. Thank!